US8244529B2 - Multi-pass echo residue detection with speech application intelligence - Google Patents
Multi-pass echo residue detection with speech application intelligence Download PDFInfo
- Publication number
- US8244529B2 US8244529B2 US13/236,968 US201113236968A US8244529B2 US 8244529 B2 US8244529 B2 US 8244529B2 US 201113236968 A US201113236968 A US 201113236968A US 8244529 B2 US8244529 B2 US 8244529B2
- Authority
- US
- United States
- Prior art keywords
- speech input
- user speech
- audio data
- predetermined
- echo residue
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 46
- 239000000284 extract Substances 0.000 claims abstract description 3
- 238000004458 analytical method Methods 0.000 claims description 21
- 238000001914 filtration Methods 0.000 claims description 19
- 238000005070 sampling Methods 0.000 claims description 4
- 238000004590 computer program Methods 0.000 claims description 2
- 230000015654 memory Effects 0.000 description 13
- 230000002596 correlated effect Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 8
- 230000007613 environmental effect Effects 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 230000000875 corresponding effect Effects 0.000 description 5
- 230000003993 interaction Effects 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 238000013461 design Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 206010011224 Cough Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000000246 remedial effect Effects 0.000 description 1
- 230000029058 respiratory gaseous exchange Effects 0.000 description 1
- 230000004043 responsiveness Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000013024 troubleshooting Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M9/00—Arrangements for interconnection not involving centralised switching
- H04M9/08—Two-way loud-speaking telephone systems with means for conditioning the signal, e.g. for suppressing echoes for one or both directions of traffic
- H04M9/082—Two-way loud-speaking telephone systems with means for conditioning the signal, e.g. for suppressing echoes for one or both directions of traffic using echo cancellers
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L21/00—Speech or voice signal processing techniques to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
- G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
- G10L21/0208—Noise filtering
- G10L2021/02082—Noise filtering the noise being echo, reverberation of the speech
Definitions
- the present invention relates to speech recognition. More particularly, the present invention relates to detection of echo residue in speech recognition systems.
- Speech recognition systems may include a speech recognition engine that recognizes speech received from a user over an incoming channel.
- the recording from the incoming channel should not contain data from the outgoing channel.
- system prompt signals should reside on the out going channel but should not carry over to the incoming channel.
- Echo residue occurs when signals on one channel (e.g., incoming) result from signals on another (e.g., outgoing) channel. Echo residue is responsible for users having poor experiences with new speech recognition systems. In particular, the echo residue on an incoming channel distorts the speech signals from the user that are to be recognized by a speech recognition system.
- Moderate echo residue can mask a user's speech as noise, and render the system non-responsive to any user input. Loud echo residue may be improperly recognized as user input, in which case a condition known as “self barge-in” occurs. There are many causes of echo residue, including loud prompts, a poor terminating device at the switch, wrong echo-cancellation settings in the telephony board, electromagnetic (EM) interference from other equipment, bad channels, bad line cards and poor speech recognition engine parameter settings. Based on the cause, the problem may be experienced consistently by all users, selectively by users on certain channels, or temporarily by users during a particular dialog state/prompt in an application.
- EM electromagnetic
- Echo residue detection is the first step to eliminating echo residue itself, particularly in situations where the echo residue is caused by factors outside of the control of the platform provider.
- echo residue in speech applications such as interactive voice response (IVR) applications may have very particular domain-specific causes.
- detection techniques may be used to isolate the causes of echo residue, and each identified cause can be individually addressed.
- FIG. 6 shows an exemplary plot portraying a recording of a conventional speech interaction on an incoming channel as amplitude versus time.
- the amplitude of the recorded signal on the plot is flat when a system prompt is playing, as the user is quietly listening and providing no input.
- the spike shown in FIG. 6 occurs when the user speaks.
- FIG. 7 shows an exemplary recording in a wave (.wav) file that contains echo residue in an incoming channel.
- a speech recognition system were capable of distinguishing when speech starts by the significantly higher amplitudes in the latter portion of the plot, it might seem that a speech recognition system could identify the echo residue by the low amplitude signals before the start of speech.
- FIG. 8 an exemplary recording that contains only normal environmental noise (e.g., cell phone static, background noise) in an incoming channel is very similar to the recording that contains echo residue as shown in FIG. 7 . Accordingly, the environmental noise has characteristics essentially identical to echo residue, and cannot be identified by signal processing techniques such as low pass filtering. As a result, a tremendous commitment of time is required for a human to manually review audio files in order to distinguish between environmental noise and echo residue.
- normal environmental noise e.g., cell phone static, background noise
- multi-pass echo residue detection with speech application intelligence is provided.
- FIG. 1 shows an exemplary general computer system that includes a set of instructions for performing a method of multi-pass echo residue detection with speech application intelligence
- FIG. 2 shows an exemplary method for multi-pass echo residue detection
- FIG. 3 shows an exemplary array of data sources that provide data to a computer system that performs a method of multi-pass echo residue detection
- FIG. 4 shows an exemplary plot of echo residue with a corresponding system prompt
- FIG. 5 shows an exemplary plot of environmental noise with a corresponding system prompt
- FIG. 6 shows an exemplary recording of a conventional speech interaction on an incoming channel
- FIG. 7 shows an exemplary recording that contains echo residue for a conventional speech interaction on an incoming channel
- FIG. 8 shows an exemplary recording that contains environmental noise for a conventional speech interaction on an incoming channel.
- a method for detecting echo residue associated with a speech application.
- the method includes correlating audio data from an input channel with audio data from an output channel to obtain a correlation result.
- the method also includes comparing a determined value of the correlation result with a predetermined threshold.
- the method additionally includes categorizing the audio data for the input channel as including an acceptable level of residual echo when the determined value of the correlation result is greater than the predetermined threshold, and categorizing the audio data for the input channel as including an unacceptable level of residual echo when the determined value of the correlation result is less than the predetermined threshold.
- the method also includes filtering audio data to determine whether the audio data should be further analyzed.
- the filtering also includes comparing a root mean square of the user input with a predetermined maximum threshold, and categorizing the user input as containing more than a predetermined level of noise when the root mean square of the user input is not less than the predetermined threshold.
- the filtering also includes comparing a root mean square of the user input with a predetermined minimum threshold, and categorizing the user input as containing less than a predetermined level of noise when the root mean square of the user input is not greater than the predetermined threshold.
- the filtering includes comparing a maximum amplitude of the user input with a predetermined threshold, and categorizing the user input as containing more than a predetermined level of noise when the maximum amplitude of the user input is not less than the predetermined threshold.
- the filtering includes determining whether the audio data contains user input.
- the correlated audio data is audio data having a level of noise within a predetermined range.
- the correlating also includes correlating user input from the input channel with a scripted audible prompt provided over the output channel.
- the filtering includes determining whether the user input is at least a predetermined duration, and extracting the predetermined length of the user input when the user input is at least the predetermined duration.
- the correlating further includes correlating the predetermined duration of the user input with a predetermined duration of the scripted audible prompt, and classifying a correlation result according to a predetermined correlation threshold.
- a computer readable medium for storing a computer program that detects echo residue associated with a speech application.
- the computer readable medium includes an audio data correlating code segment that correlates audio data from an input channel with audio data from an output channel to obtain a correlation result.
- the computer readable medium also includes a comparing code segment that compares a determined value of the correlation result with a predetermined threshold.
- the computer readable medium further includes a categorizing code segment that categorizes the audio data for the input channel as including an acceptable level of residual echo when the determined value of the correlation result is greater than the predetermined threshold, and that categorizes the audio data for the input channel as including an unacceptable level of residual echo when the determined value of the correlation result is less than the predetermined threshold.
- the computer readable medium also includes a filtering code segment that filters audio data to determine whether the audio data should be further analyzed.
- the filtering code segment compares a root mean square of the user input with a predetermined maximum threshold, and categorizes the user input as containing more than a predetermined level of noise when the root mean square of the user input is not less than the predetermined threshold.
- the filtering code segment compares a root mean square of the user input with a predetermined minimum threshold, and categorizes the user input as containing less than a predetermined level of noise when the root mean square of the user input is not more than the predetermined threshold.
- the filtering code segment compares a maximum amplitude of the user input with a predetermined threshold, and categorizes the user input as containing more than a predetermined level of noise when the maximum amplitude of the user input is not less than the predetermined threshold.
- the filtering code segment determines whether the audio data contains user input.
- the correlated audio data is audio data having a level of noise within a predetermined range.
- the correlating code segment correlates user input from the input channel with a scripted audible prompt provided over the output channel.
- the filtering code segment determines whether the user input is at least a predetermined duration, and extracts the predetermined length of the user input when the user input is at least the predetermined duration.
- the correlating code segment correlates the predetermined duration of the user input with a predetermined length time of the scripted audible prompt, and classifies a correlation result according to a predetermined correlation threshold.
- an echo residue detector associated with a speech application includes an input port through which audio data from an input channel is received.
- the echo residue detector also includes an output port through which audio data from an output channel is transmitted.
- the echo residue detector further includes a processor that correlates the audio data from the input channel with the audio data from the output channel to obtain a correlation result.
- a determined value of the correlation result is compared with a predetermined threshold.
- the audio data for the input channel is categorized as including an acceptable level of residual echo when the determined value of the correlation result is greater than the predetermined threshold, and the audio data for the input channel is categorized as including an unacceptable level of residual echo when the determined value of the correlation result is less than the predetermined threshold.
- the processor filters audio data to determine whether the audio data should be further analyzed.
- the filtering includes comparing a root mean square of the user input with a predetermined maximum threshold, and categorizing the user input as containing more than a predetermined level of noise when the root mean square of the user input is not less than the predetermined threshold.
- the filtering includes comparing a root mean square of the user input with a predetermined minimum threshold, and categorizing the user input as containing less than a predetermined level of noise when the root mean square of the user input is not greater than the predetermined threshold.
- the filtering includes comparing a maximum amplitude of the user input with a predetermined threshold, and categorizing the user input as containing more than a predetermined level of noise when the maximum amplitude of the user input is not less than the predetermined threshold.
- the filtering includes determining whether the audio data contains user input.
- the correlated audio data is audio data having a level of noise within a predetermined range.
- user input from the input channel is correlated with a scripted audible prompt provided over the output channel.
- the present invention leverages speech domain-specific techniques to detect speech application echo residue. As described herein, simple analysis and on-board cancellation concepts are used for fast, accurate and automated echo residue detection. Because echo residue detection is needed before one can eliminate echo residue, the present invention can be used as a basis for taking remedial steps to determine causes for echo residue, and for taking steps to eliminate the causes of echo residue.
- FIG. 1 an illustrative embodiment of a general computer system, on which multi-pass echo residue detection with speech application intelligence can be implemented, is shown and is designated 100 .
- the computer system 100 can include a set of instructions that can be executed to cause the computer system 100 to perform any one or more of the methods or computer based functions disclosed herein.
- the computer system 100 may operate as a standalone device or may be connected, e.g., using a network 101 , to other computer systems or peripheral devices.
- the computer system may operate in the capacity of a server or as a client user computer in a server-client user network environment, or as a peer computer system in a peer-to-peer (or distributed) network environment.
- the computer system 100 can also be implemented as or incorporated into various devices, such as a personal computer (PC), a tablet PC, a set-top box (STB), a personal digital assistant (PDA), a mobile device, a palmtop computer, a laptop computer, a desktop computer, a communications device, a wireless telephone, a land-line telephone, a control system, a camera, a scanner, a facsimile machine, a printer, a pager, a personal trusted device, a web appliance, a network router, switch or bridge, or any other machine capable of executing a set of instructions (sequential or otherwise) that specify actions to be taken by that machine.
- the computer system 100 can be implemented using electronic devices that provide voice, video or data communication.
- the term “system” shall also be taken to include any collection of systems or sub-systems that individually or jointly execute a set, or multiple sets, of instructions to perform one or more computer functions.
- the computer system 100 may include a processor 110 , e.g., a central processing unit (CPU), a graphics processing unit (GPU), or both. Moreover, the computer system 100 can include a main memory 120 and a static memory 130 that can communicate with each other via a bus 108 . As shown, the computer system 100 may further include a video display unit 150 , such as a liquid crystal display (LCD), an organic light emitting diode (OLED), a flat panel display, a solid state display, or a cathode ray tube (CRT). Additionally, the computer system 100 may include an input device 160 , such as a keyboard, and a cursor control device 170 , such as a mouse. The computer system 100 can also include a disk drive unit 180 , a signal generation device 190 , such as a speaker or remote control, and a network interface device 140 .
- a processor 110 e.g., a central processing unit (CPU), a graphics processing unit (GPU), or both.
- the disk drive unit 180 may include a computer-readable medium 182 in which one or more sets of instructions 184 , e.g. software, can be embedded. Further, the instructions 184 may embody one or more of the methods or logic as described herein. In a particular embodiment, the instructions 184 may reside completely, or at least partially, within the main memory 120 , the static memory 130 , and/or within the processor 110 during execution by the computer system 100 . The main memory 120 and the processor 110 also may include computer-readable media.
- dedicated hardware implementations such as application specific integrated circuits, programmable logic arrays and other hardware devices, can be constructed to implement one or more of the methods described herein.
- Applications that may include the apparatus and systems of various embodiments can broadly include a variety of electronic and computer systems.
- One or more embodiments described herein may implement functions using two or more specific interconnected hardware modules or devices with related control and data signals that can be communicated between and through the modules, or as portions of an application-specific integrated circuit. Accordingly, the present system encompasses software, firmware, and hardware implementations.
- the methods described herein may be implemented by software programs executable by a computer system.
- implementations can include distributed processing, component/object distributed processing, and parallel processing.
- virtual computer system processing can be constructed to implement one or more of the methods or functionality as described herein.
- the present disclosure contemplates a computer-readable medium 182 that includes instructions 184 or receives and executes instructions 184 responsive to a propagated signal, so that a device connected to a network 101 can communicate voice, video or data over the network 101 . Further, the instructions 184 may be transmitted or received over the network 101 via the network interface device 140 .
- While the computer-readable medium is shown to be a single medium, the term “computer-readable medium” includes a single medium or multiple media, such as a centralized or distributed database, and/or associated caches and servers that store one or more sets of instructions.
- the term “computer-readable medium” shall also include any medium that is capable of storing, encoding or carrying a set of instructions for execution by a processor or that cause a computer system to perform any one or more of the methods or operations disclosed herein.
- the computer-readable medium can include a solid-state memory such as a memory card or other package that houses one or more non-volatile read-only memories. Further, the computer-readable medium can be a random access memory or other volatile re-writable memory. Additionally, the computer-readable medium can include a magneto-optical or optical medium, such as a disk or tapes or other storage device to capture carrier wave signals such as a signal communicated over a transmission medium. A digital file attachment to an e-mail or other self-contained information archive or set of archives may be considered a distribution medium that is equivalent to a tangible storage medium. Accordingly, the disclosure is considered to include any one or more of a computer-readable medium or a distribution medium and other equivalents and successor media, in which data or instructions may be stored.
- multiple filters may be used to determine which audio data is most likely to include acceptable levels of residual echo, which audio data is most likely not to include acceptable levels of residual echo, and which audio data is unsuitable for analysis.
- Audio data from an incoming channel that passes through the filters can be correlated with audio data from an outgoing channel to perform the actual detection of residual echo.
- multi-pass echo residue detection with speech application intelligence can avoid committing the significant computing resources that would otherwise be necessary to correlate all incoming audio data for a speech application.
- the use of filters is capable of effectively eliminating 80% or more of input wave files without requiring the correlation of signals from the incoming channel and the outgoing channel.
- FIG. 2 shows an exemplary method for multi-pass echo residue detection. To be exact, FIG. 2 shows an exemplary multi-pass echo residue detection algorithm. Numeric values used in the algorithm are for illustration purposes, and values used in operation may be adjusted as appropriate.
- information from a recognition log or other information repository is fed to a software component which analyzes the information and determines that audio data is present.
- the information from the speech recognition engine log includes a wave (.wav) file of audio data.
- the analysis of information from the recognition engine log may occur in real-time, or may be performed periodically on a batch of data.
- a determination is made whether the audio input contains user input. If the audio input contains user input (S 202 Yes), the time where speech started is extracted from the speech recognition engine log at S 204 .
- the wave file contains speech
- only the pre-“start of speech” section should be used for comparison with the system prompt.
- the input file should be at least the duration of the system prompt.
- system speech prompts are designed to last at least 6 seconds. Accordingly, the wave file of audio data should contain at least 6 seconds of speech from a speech prompt, even if no user speech is contained in the audio input.
- the pre-“start of speech” duration needs to have a minimum duration of, e.g., at least 4 seconds for wave files at an 8000 KHz sampling rate, for the echo residue detection to be reliable. If the pre-“start of speech” duration is not the minimum duration, e.g., at least 4 seconds for wave files at an 8000 KHz sampling rate, the risk of false accepts is unacceptable due to sampling error and similarity of different speech patterns having similar numbers of syllables. However, correlation beyond a certain data length provides little return on accuracy, at the cost of computational time. For the embodiment shown in FIG. 2 , 6 seconds is used as an optimal cut off duration. Since speech less than 6 seconds yields poor analysis results, the analysis ends.
- a determination is made at S 208 whether the time when speech started is less than a predetermined threshold of 6 seconds. If the time when speech started is less than 6 seconds (S 208 Yes), the analysis ends.
- the first 6 seconds of the audio data is extracted at S 210 .
- the next set of filters shown in FIG. 2 determine that the first 6 seconds contains a low level of noise.
- root mean square (RMS) is used as a first indicator of consistent noise.
- RMS root mean square
- noise is judged to be tolerable.
- the level of echo is determined to be acceptable. No on-board echo-canceling system is perfect, so allowing for a negligible level of echo would avoid the analysis to be over-sensitive.
- the system checks for maximum amplitude, to eliminate files with speech that is early and soft, speech that the recognition engine fails to recognize.
- the audio data will be cross-correlated. Since the speech prompts are consistent, the speed of analysis can be increased by caching the first 6 seconds from the prompt wave files to further increase speed. In an analysis that involves tens of thousands of wave files, for an application that only has a few dozen dialog states, the difference in performance may be very significant.
- the system prompt which corresponds to the user audio data is determined.
- the input audio data from the speech recognition engine log is accompanied by identifying information such as source, channel and system prompt.
- the identifying information may also include an identification of an interactive speech recognition platform that was used to interact with the caller who experienced echo residue.
- FIG. 3 shows an exemplary array of data sources that provide data to and/or receive data from a computer system 390 such as the computer system shown in FIG. 1 .
- a computer system that performs the echo residue detection may receive data from a recognition engine log 310 , incoming wave files 320 , a telephone log 330 , a dialogue design tool 340 , trimmed system prompts 350 in a memory and/or a database of system prompts 360 .
- the computer system 390 stores detection results and identifying information in a data warehouse 370 .
- FIGS. 4 and 5 show exemplary plots of audio data that will result in different outcomes according to the method shown in FIG. 2 .
- FIG. 4 shows an exemplary plot of echo residue with a corresponding system prompt.
- the wave files played by the system are predetermined, and either documented in a dialog design tool as shown in FIG. 3 , or captured in a real-time telephony log.
- a dialog design tool documents the dialog between the user and the speech recognition system, and specifies which prompts to play following each user response.
- a real-time telephony log contains the actual wave file name that the telephony platform played to the user.
- a system prompt is correlated with a low volume signal from the input channel, so that the detection of echo residue is automated.
- the echo detection program examines the peak value of the correlation result at S 228 , and determines whether a high correlation exists between the input and output signals. In FIG. 4 , a prominent peak is clearly shown. The prominent peak demonstrates the two signals are highly correlated so that the audio data in FIG. 4 would be classified as having unacceptable residual echo.
- FIG. 5 shows an exemplary plot of environmental noise with a corresponding system prompt.
- the result is completely different from the case of echo residue shown in FIG. 4 .
- the plot of noise crossing includes a much lower peak value, and the audio data shown in FIG. 5 would be classified as having acceptable residual echo.
- the detection of residual echo can be automated for speech applications such as speech recognition engines.
- user audio input data can be classified as having either an acceptable or unacceptable level of residual echo.
- user audio input data can be parsed with one or more filters so that large amounts of correlation processing can be avoided for user audio input data that is not likely to produce reliable results. This optimization of the analysis results can be used to isolate the strongest examples of residual echo, such that the causes of the residual echo can be identified and remedied.
- inventions of the disclosure may be referred to herein, individually and/or collectively, by the term “invention” merely for convenience and without intending to voluntarily limit the scope of this application to any particular invention or inventive concept.
- inventions merely for convenience and without intending to voluntarily limit the scope of this application to any particular invention or inventive concept.
- specific embodiments have been illustrated and described herein, it should be appreciated that any subsequent arrangement designed to achieve the same or similar purpose may be substituted for the specific embodiments shown.
- This disclosure is intended to cover any and all subsequent adaptations or variations of various embodiments. Combinations of the above embodiments, and other embodiments not specifically described herein, will be apparent to those of skill in the art upon reviewing the description.
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Telephone Function (AREA)
Abstract
Description
Claims (20)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/236,968 US8244529B2 (en) | 2005-09-12 | 2011-09-20 | Multi-pass echo residue detection with speech application intelligence |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US11/223,119 US8041564B2 (en) | 2005-09-12 | 2005-09-12 | Multi-pass echo residue detection with speech application intelligence |
US13/236,968 US8244529B2 (en) | 2005-09-12 | 2011-09-20 | Multi-pass echo residue detection with speech application intelligence |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/223,119 Continuation US8041564B2 (en) | 2005-09-12 | 2005-09-12 | Multi-pass echo residue detection with speech application intelligence |
Publications (2)
Publication Number | Publication Date |
---|---|
US20120008752A1 US20120008752A1 (en) | 2012-01-12 |
US8244529B2 true US8244529B2 (en) | 2012-08-14 |
Family
ID=37856395
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/223,119 Expired - Fee Related US8041564B2 (en) | 2005-09-12 | 2005-09-12 | Multi-pass echo residue detection with speech application intelligence |
US13/236,968 Active US8244529B2 (en) | 2005-09-12 | 2011-09-20 | Multi-pass echo residue detection with speech application intelligence |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/223,119 Expired - Fee Related US8041564B2 (en) | 2005-09-12 | 2005-09-12 | Multi-pass echo residue detection with speech application intelligence |
Country Status (2)
Country | Link |
---|---|
US (2) | US8041564B2 (en) |
CA (1) | CA2557977A1 (en) |
Families Citing this family (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101617363B (en) * | 2007-02-21 | 2012-09-05 | 艾利森电话股份有限公司 | Double talk detector |
JP5156043B2 (en) * | 2010-03-26 | 2013-03-06 | 株式会社東芝 | Voice discrimination device |
US9779731B1 (en) * | 2012-08-20 | 2017-10-03 | Amazon Technologies, Inc. | Echo cancellation based on shared reference signals |
US10147441B1 (en) | 2013-12-19 | 2018-12-04 | Amazon Technologies, Inc. | Voice controlled system |
US20180196919A1 (en) * | 2017-01-10 | 2018-07-12 | International Business Machines Corporation | Automated health dialoguing and action enhancement |
CN110148421B (en) * | 2019-06-10 | 2021-07-20 | 浙江大华技术股份有限公司 | Residual echo detection method, terminal and device |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5668865A (en) | 1996-02-26 | 1997-09-16 | Lucent Technologies Inc. | Echo canceler E-side speech detector |
US6035033A (en) | 1996-09-26 | 2000-03-07 | Siemens Aktiengesellschaft | Method and apparatus for limiting residual echo in a speech signal-carrying channel or line |
US6044108A (en) | 1997-05-28 | 2000-03-28 | Data Race, Inc. | System and method for suppressing far end echo of voice encoded speech |
US6178162B1 (en) | 1997-12-19 | 2001-01-23 | Nortel Networks Corporation | Method and apparatus for inhibiting echo in a channel of a communication system |
US6434234B1 (en) | 1998-10-31 | 2002-08-13 | Alcatel | Process for improving the echo suppression in a telecommunications system |
US6449361B1 (en) | 1998-03-31 | 2002-09-10 | Sanyo Electric Co., Ltd. | Control method and device for echo canceller |
US20020169602A1 (en) | 2001-05-09 | 2002-11-14 | Octiv, Inc. | Echo suppression and speech detection techniques for telephony applications |
US6570986B1 (en) | 1999-08-30 | 2003-05-27 | Industrial Technology Research Institute | Double-talk detector |
US6744884B1 (en) | 2000-09-29 | 2004-06-01 | Skyworks Solutions, Inc. | Speaker-phone system change and double-talk detection |
US6744885B1 (en) | 2000-02-24 | 2004-06-01 | Lucent Technologies Inc. | ASR talkoff suppressor |
US6766019B1 (en) | 2000-07-21 | 2004-07-20 | Agere Systems Inc. | Method and apparatus for performing double-talk detection in acoustic echo cancellation |
US6804203B1 (en) | 2000-09-15 | 2004-10-12 | Mindspeed Technologies, Inc. | Double talk detector for echo cancellation in a speech communication system |
US6873704B1 (en) | 1998-10-13 | 2005-03-29 | Samsung Electronics Co., Ltd | Apparatus for removing echo from speech signals with variable rate |
US6944288B2 (en) | 2001-01-11 | 2005-09-13 | Zarlink Semiconductor Inc. | Double-talk and path change detection using a matrix of correlation coefficients |
US20070055522A1 (en) | 2005-08-26 | 2007-03-08 | Sbc Knowledge Ventures, L.P. | Self-learning multi-source speech data reconstruction |
-
2005
- 2005-09-12 US US11/223,119 patent/US8041564B2/en not_active Expired - Fee Related
-
2006
- 2006-08-22 CA CA002557977A patent/CA2557977A1/en not_active Abandoned
-
2011
- 2011-09-20 US US13/236,968 patent/US8244529B2/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5668865A (en) | 1996-02-26 | 1997-09-16 | Lucent Technologies Inc. | Echo canceler E-side speech detector |
US6035033A (en) | 1996-09-26 | 2000-03-07 | Siemens Aktiengesellschaft | Method and apparatus for limiting residual echo in a speech signal-carrying channel or line |
US6044108A (en) | 1997-05-28 | 2000-03-28 | Data Race, Inc. | System and method for suppressing far end echo of voice encoded speech |
US6178162B1 (en) | 1997-12-19 | 2001-01-23 | Nortel Networks Corporation | Method and apparatus for inhibiting echo in a channel of a communication system |
US6449361B1 (en) | 1998-03-31 | 2002-09-10 | Sanyo Electric Co., Ltd. | Control method and device for echo canceller |
US6873704B1 (en) | 1998-10-13 | 2005-03-29 | Samsung Electronics Co., Ltd | Apparatus for removing echo from speech signals with variable rate |
US6434234B1 (en) | 1998-10-31 | 2002-08-13 | Alcatel | Process for improving the echo suppression in a telecommunications system |
US6570986B1 (en) | 1999-08-30 | 2003-05-27 | Industrial Technology Research Institute | Double-talk detector |
US6744885B1 (en) | 2000-02-24 | 2004-06-01 | Lucent Technologies Inc. | ASR talkoff suppressor |
US6766019B1 (en) | 2000-07-21 | 2004-07-20 | Agere Systems Inc. | Method and apparatus for performing double-talk detection in acoustic echo cancellation |
US6804203B1 (en) | 2000-09-15 | 2004-10-12 | Mindspeed Technologies, Inc. | Double talk detector for echo cancellation in a speech communication system |
US6744884B1 (en) | 2000-09-29 | 2004-06-01 | Skyworks Solutions, Inc. | Speaker-phone system change and double-talk detection |
US6944288B2 (en) | 2001-01-11 | 2005-09-13 | Zarlink Semiconductor Inc. | Double-talk and path change detection using a matrix of correlation coefficients |
US20020169602A1 (en) | 2001-05-09 | 2002-11-14 | Octiv, Inc. | Echo suppression and speech detection techniques for telephony applications |
US20070055522A1 (en) | 2005-08-26 | 2007-03-08 | Sbc Knowledge Ventures, L.P. | Self-learning multi-source speech data reconstruction |
Also Published As
Publication number | Publication date |
---|---|
CA2557977A1 (en) | 2007-03-12 |
US20120008752A1 (en) | 2012-01-12 |
US20070061134A1 (en) | 2007-03-15 |
US8041564B2 (en) | 2011-10-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US8244529B2 (en) | Multi-pass echo residue detection with speech application intelligence | |
US8417524B2 (en) | Analysis of the temporal evolution of emotions in an audio interaction in a service delivery environment | |
US9258425B2 (en) | Method and system for speaker verification | |
US8798255B2 (en) | Methods and apparatus for deep interaction analysis | |
US8219404B2 (en) | Method and apparatus for recognizing a speaker in lawful interception systems | |
US8065146B2 (en) | Detecting an answering machine using speech recognition | |
US20070203708A1 (en) | System and method for providing transcription services using a speech server in an interactive voice response system | |
US8929519B2 (en) | Analyzing speech application performance | |
US9583108B2 (en) | Voice detection for automated communication system | |
EP1936607B1 (en) | Automated speech recognition application testing | |
CN111128241A (en) | Intelligent quality inspection method and system for voice call | |
US9406075B1 (en) | System and method for improving tuning using user provided satisfaction scores | |
US8130916B2 (en) | Dynamically improving performance of an interactive voice response (IVR) system using a complex events processor (CEP) | |
US8606585B2 (en) | Automatic detection of audio advertisements | |
US8949134B2 (en) | Method and apparatus for recording/replaying application execution with recorded voice recognition utterances | |
US9047872B1 (en) | Automatic speech recognition tuning management | |
CN112087726B (en) | Method and system for identifying polyphonic ringtone, electronic equipment and storage medium | |
KR20170010978A (en) | Method and apparatus for preventing voice phishing using pattern analysis of communication content | |
WO2021143095A1 (en) | Dialing test method and apparatus, and computer device and storage medium | |
US11450336B1 (en) | System and method for smart feedback cancellation | |
CN111049997B (en) | Telephone background music detection model method, system, equipment and medium | |
CN112487144A (en) | Multi-round conversation robot intention hit optimization method and device | |
CN111863004A (en) | Sound signal processing method, system, electronic device and storage medium | |
CN118018649A (en) | Outbound efficiency improving method, system and storage medium based on artificial intelligence | |
EP2608506A1 (en) | A System And Method For Developing and monitoring IVR dialogs |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: SBC KNOWLEDGE VENTURES, L.P., NEVADA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:WONG, NGAI CHIU;REEL/FRAME:026935/0344 Effective date: 20051012 |
|
FEPP | Fee payment procedure |
Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
FPAY | Fee payment |
Year of fee payment: 4 |
|
AS | Assignment |
Owner name: AT&T KNOWLEDGE VENTURES, L.P., NEVADA Free format text: CHANGE OF NAME;ASSIGNOR:SBC KNOWLEDGE VENTURES, L.P.;REEL/FRAME:039096/0550 Effective date: 20060224 Owner name: AT&T INTELLECTUAL PROPERTY I, L.P., GEORGIA Free format text: CHANGE OF NAME;ASSIGNOR:AT&T KNOWLEDGE VENTURES, L.P.;REEL/FRAME:039097/0003 Effective date: 20071001 |
|
AS | Assignment |
Owner name: NUANCE COMMUNICATIONS, INC., MASSACHUSETTS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:AT&T INTELLECTUAL PROPERTY I, L.P.;REEL/FRAME:041504/0952 Effective date: 20161214 |
|
MAFP | Maintenance fee payment |
Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Year of fee payment: 8 |
|
FEPP | Fee payment procedure |
Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |